Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: xtmixed and time varying level 1 fixed effects

From   Christoph Wunder <>
Subject   st: xtmixed and time varying level 1 fixed effects
Date   Thu, 22 Sep 2005 14:42:26 +0200

Dear Statalist,

I am dealing with a two level random intercept model. I use panel data
with (repeated) observations nested in individuals (level 2) nested in
households (level 1). Consequently, the model can be specified as

y_hit = x_hit * b0 + n_hit

where y_hit denotes the observation for individual i in household h at
time t; n is the nested error component (n_hit = u_h + v_hi + e_hit). I
estimate this model using

xtmixed y x || level1: || level2:

So far, the procedure seems to be straight forward. In a second step, however, time varying and household specific variables enter the model.

y_hit = x_hit * b0 + z_ht * b1 + n_hit

Question: How can I control for the fact that individuals living in the
same household share the same household characteristics at time t? Or
can z_ht be treated as a household fixed effect (i.e. is the estimation
identical to a model containing z_h)?

* For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index